Nutrients

Vitamin D

Vitamin D is not really a vitamin. It is a hormone precursor, and most people get very little of it from food. Ultraviolet B light converts a cholesterol derivative in the skin into vitamin D, the liver turns that into the storage form, and the kidney turns the storage form into the active hormone. That last step is why the Kidneys section mentions vitamin D, since a kidney under strain is also a kidney activating less of it.

Melanin is the first thing that limits that. It is the pigment that gives skin its colour, and it absorbs ultraviolet light across the same band the skin needs for the conversion. Melanin and the vitamin D precursor are competing for the same photons, so the more melanin there is, the fewer of them reach the molecule that has to be changed. The same hour outdoors produces less vitamin D in darker skin than in lighter skin. That much is settled.

How much less is not settled. The figure you will see almost everywhere, that darker skin needs about 6 times as long in the sun, traces to a letter published in 1982 describing 5 volunteers, in which a single dark-skinned volunteer was re-exposed at 6 times the dose and came out level with the lighter-skinned ones. The measured UK figure is a good deal smaller than that, and it is set out in the Vitamin D entry in the Biomarker Guide.

The receptor it acts on is present in most tissues in the body, not only bone and gut, which is the origin of the very long list of things vitamin D has been investigated for. Presence of a receptor is not the same as demonstrated benefit, and that gap is where most of the overclaiming in this area lives.

What we measure is the storage form, 25-hydroxyvitamin D, rather than the active hormone. That is deliberate and standard everywhere. The storage form reflects supply, has a long half-life and is stable. The active hormone is tightly regulated and can sit at a perfectly normal level while stores are empty.

An open question underneath that measurement

Around 85% to 90% of the vitamin D in your blood is locked to a carrier protein and is not available to tissue. Roughly another 10% to 15% rides loosely on albumin, and less than 1% is genuinely free. The same logic that applies to testosterone and SHBG applies here, and what a tissue can use is the unbound part rather than the total.

How much carrier protein a person makes varies, partly for inherited reasons, which means two people with identical total vitamin D can in principle have quite different amounts available. Some studies find the free or available fraction tracks parathyroid hormone, calcium and bone density better than the total does. Others find it adds nothing. The assays are difficult and disagree with each other, and we are not aware of a guideline that uses it.

So this is genuinely unsettled rather than settled-but-ignored. Know that the question exists rather than assuming the total is the whole answer.

The numbers, and why they are argued about

In the UK there is one figure with real authority behind it, and a band underneath it with rather less. The Scientific Advisory Committee on Nutrition sets 25 nmol/L as a population protection threshold, below which the risk of poor musculoskeletal health rises measurably. That is the figure the National Institute for Health and Care Excellence adopts.

And it is not what most people assume it is. It was set as the level people in Britain should stay above year-round to avoid skeletal disease, and the committee that set it cautioned against using it as a diagnostic cut-off for an individual. It has been used as one anyway.

The band most labs print between 25 and 50 and label insufficiency is a clinical and lab convention rather than a position of that committee. It is widely used and it is not unreasonable. It simply does not carry the same authority as the floor beneath it, and we are not aware of any UK body publishing an optimal level above that floor.

The floor itself is also actively contested. A UK position statement published in 2026 argues the deficiency threshold should be raised to 50 nmol/L, on the grounds that the current figure understates the scale of the problem and sits below most international practice.

And in one significant place the argument has recently gone the other way. The 75 nmol/L figure that circulates everywhere as the level to aim for came from the Endocrine Society in 2011. In 2024 the same body reviewed the evidence again, concluded there was no clear evidence defining an optimal target, withdrew the figure, and stopped treating insufficiency as a category worth acting on at all. It also came out against routine vitamin D testing in healthy adults with no medical reason for it.

Notice which way that went. Guidance usually lags the evidence rather than reversing itself. Here a body that had set a high target looked again and took it back down, and almost every website still quotes the old number.

So the position, and it is an inference rather than a guideline figure, is that within reason higher is better than lower across the lower half of that range, and the disagreement is about where "adequate" starts rather than about whether deficiency matters. There is also a ceiling, covered below, so this is not an argument that more is always better.

Two things specific to living here

The UK has a latitude problem. Between roughly October and March the sun does not climb high enough for skin synthesis to happen at all, anywhere in Britain. That is about the angle the light arrives at rather than how cold or bright it feels. Low sun means a longer path through the atmosphere, and the ultraviolet B band is filtered out of it almost entirely while the visible light that makes a day look bright still gets through. A clear cold January afternoon delivers none of the wavelength that does the work, and a mild February day delivers none of it either. Levels fall through autumn and winter for almost everyone, which is why a result taken in one season is not comparable with one taken in another.

And this is the one nutrient here where too much is a realistic risk. It comes from supplements rather than from sun or food, because the skin simply stops producing once it has enough. Excess vitamin D raises calcium, and that is the route to nausea, thirst, kidney stones and, over time, kidney damage. High-dose supplementation without testing is the usual way people get there.

How much skin is uncovered

Latitude sets the ceiling. What is left uncovered under that ceiling is the other half of it, and in the UK that has been measured rather than assumed. Shorts and a t-shirt leave about 23% of body surface exposed. A long-sleeved top, trousers or a long skirt, and a head covering leave about 4%.

That is clothing and what people do at weekends rather than anything inherited. The distinction matters, because behaviour can change and genes cannot. What the measured differences between groups look like is in the Vitamin D entry in the Biomarker Guide.

What food adds

Food is the other route in, and it carries far less than summer skin does. Oily fish is the main natural source, with eggs and fortified spreads and cereals behind it. Measured intakes across UK South Asian adults run at roughly 1 to 3 micrograms a day, which is a fraction of what a good summer produces, and differences between groups in how much oily fish gets eaten are large enough to show up in their results.

B12 and folate

These two are measured together because they do overlapping work, and because one of them can conceal a deficiency in the other. The mechanism behind that explains almost everything else about how these markers behave.

Both are required to build DNA, which matters most in tissue that divides rapidly, and bone marrow above all. When either runs short, red cells are produced fewer and larger than they should be. That is why a deficiency in either shows up in the full blood count as a raised MCV, and why the Blood cells section treats cell size as a pointer rather than a curiosity.

B12 takes a more complicated route into the body than any other nutrient. It is handed between three different carrier proteins on the way. One protects it through the acid of the stomach. A second, made by the stomach lining and called intrinsic factor, escorts it to a specific receptor far down the small intestine, which is the only place it can be absorbed. A third carries it in the blood and is the only one that delivers it into cells.

Two consequences follow. Deficiency usually fails at absorption rather than at intake. Autoimmune damage to the stomach lining, which is pernicious anaemia, or surgery and disease affecting the far end of the small intestine, will do it on a perfectly adequate diet. And most of the B12 in your blood is riding on the wrong carrier, since the majority is bound to the blood protein that does not deliver to cells, so a total B12 can read comfortably normal while the usable fraction is low. Active B12 measures the fraction on the delivering carrier, which is why it falls earlier.

Intake does matter, for a smaller group than people assume. B12 is made by bacteria rather than by plants, and it reaches you almost entirely through animal foods. Meat, fish, eggs and dairy carry it, plants do not, and beyond those it comes only from fortified foods and supplements. A diet with no animal food in it is the situation where intake rather than absorption is what fails.

That makes the usual shorthand about South Asian diets wrong. Vegetarianism among South Asian people in the UK is almost entirely an Indian pattern. In the UK data, 27% of the Indian group were vegetarian against under 1% of the Pakistani and Bangladeshi groups. Treating South Asian as a synonym for vegetarian is wrong for 2 of the 3 largest groups, and right for only about a quarter of the third.

And the assumption about ethnicity runs the other way from the evidence. The largest UK set of B12 results shows no difference between Asian and White patients. What does differ is that Black patients run higher. So what matters here is the diet rather than the ethnicity, and the diet is a minority position even inside the group where it is common.

Inside the cell, B12 does exactly two jobs, and that is the key to the rest.

The first is helping an enzyme convert homocysteine into methionine, using folate as the source of the chemical group being transferred. The second, entirely separate, is helping a different enzyme process a fragment left over from breaking down certain fats and amino acids.

When B12 runs short, both stall. The material each enzyme was meant to process backs up. Homocysteine from the first, methylmalonic acid from the second. That is why those 2 substances are the functional tests for B12 deficiency, because measuring the consequence is more reliable than measuring the vitamin.

In UK practice they are reached for in a specific situation rather than routinely. Front-line testing is a total or active B12, and where a total B12 comes back in the indeterminate band alongside symptoms, methylmalonic acid is the further test NICE points to. It narrows the question rather than closing it, and NICE says in the guideline itself that there is no gold standard test for B12 deficiency.

Between the two, UK guidance points to methylmalonic acid, and the reason is specificity rather than sensitivity. Homocysteine is the more sensitive of the pair and rises early, sometimes before any symptoms appear. But it also goes up with folate deficiency, with B6 deficiency, with reduced kidney function and with certain medicines, so a raised result hands you a list rather than an answer. Methylmalonic acid has false positives of its own, mainly kidney disease and bacterial overgrowth in the small bowel, but a markedly raised one is almost always B12.

And this is the folate trap. Folate arrives at that first enzyme carrying the group it is about to donate, and can only be converted back into the form used for DNA synthesis by donating it. If B12 is missing, the enzyme cannot run, so the folate cannot unload, and it accumulates stuck in a form the cell cannot use. The result is a functional folate deficiency inside cells that have plenty of folate in them. That is why B12 deficiency produces a blood picture identical to folate deficiency. At the level of the marrow, it is folate deficiency.

Give folate in that situation and enough of it enters by a route that bypasses the blockage to restart DNA synthesis. The red cells return to normal size and the anaemia corrects. But the second B12 job, the one that has nothing to do with folate, is untouched. And that is the one tied to nerve damage. So the finding that would have prompted investigation is erased while the underlying deficiency is still there.

One part of that story is settled and one part is not, and they are usually told as though both were settled. That folate corrects the blood picture and removes the clue is not in dispute. The stronger claim, that folate actively accelerates the nerve damage, rests largely on case reports from the 1940s, when pernicious anaemia was treated with high-dose folic acid because B12 was not yet available, and a historical re-examination has argued those cases do not support it. What is more recent, and more careful, is the observation that people with low B12 and high folate have higher homocysteine and methylmalonic acid and score worse on cognitive testing than people with low B12 and ordinary folate. That is correlation rather than proof, and it is enough to justify checking B12 before treating folate, which is what guidance does.

Serum folate, separately, reflects recent intake rather than long-term stores. It has a shorter memory than most things on a blood panel.

Folic acid will be added to flour from 13 December 2026, which will lift the amount of folate in the everyday food supply. The requirement covers non-wholemeal wheat flour, at 0.25 milligrams per 100 grams. Wholemeal flour is exempt, and so are rice flour, maize flour and the output of the smallest mills.

That exemption matters here. Chapati atta is characteristically wholemeal, so a household where most of the flour eaten is atta may see less of the added folic acid than a household eating mostly white bread. White flour used for naan, paratha and pastry is covered, and many flours sold as chapati flour are blends.

B12 deficiency causes two problems that do not arrive together. The blood picture is one. Nerve damage is the other, meaning numbness, pins and needles and problems with balance, and it can appear before the blood count changes at all, and is not always reversible once established. That asymmetry, plus the trap above, is why the two are always read together rather than separately.

NICE published a guideline on B12 deficiency in 2024 and put that point in writing. Do not rule out a B12 deficiency on the absence of anaemia or a raised cell size alone. It is an unusually blunt instruction for a guideline, and it exists because the opposite happens constantly. The same document lists difficulty concentrating and short-term memory trouble, the thing people call brain fog, among the reasons to test in the first place.

It also quietly dropped the phrase "pernicious anaemia", on the grounds that anaemia severe enough to earn the name is now rare and that the term hides what is actually happening, which is an autoimmune attack on the stomach lining. The old name is used above because it is the one people recognise, and the two mean the same thing.

One cause needs naming because it defeats the standard test. Nitrous oxide, used recreationally, chemically inactivates B12 by altering the cobalt atom at its centre. The vitamin is still present and still measures normally. It simply no longer works. Which is why guidance is explicit that where nitrous oxide is the suspected cause, the first test should be homocysteine or methylmalonic acid rather than a B12 measurement at all.

We offer neither. Methylmalonic acid is the more practical of the two outside hospital; homocysteine has a sample-handling requirement that rules out posting it entirely, for the same reason potassium is not on our menu. Saying so is better than selling the wrong one.

Tablets or injections, and why the answer has changed

Anyone diagnosed with a B12 deficiency caused by an absorption problem has usually been told the same thing. Injections, for life. That was the standard answer for decades and the logic was sound. If the escort protein is missing, the receptor at the far end of the small intestine cannot work, so swallowing B12 is pointless.

Except that roughly 1% of an oral dose is absorbed anyway, by simple diffusion straight across the gut wall, with no escort and no receptor involved at all. That 1% of a normal dose is nothing. The same 1% of a very large dose is a working treatment. That is the entire basis for high-dose oral B12, and it is why it can work in people whose absorption machinery is broken.

A Cochrane review found high-dose oral B12 corrected blood levels about as well as injections but it did not mention improvement in clinical signs and symptoms. NICE moved on this in 2024, and where the cause is dietary, or uncertain with no malabsorption suspected, oral replacement is now the first thing to consider rather than the fallback. Injections are still what NICE offers where the stomach or the far end of the small intestine has been removed. Also where another condition could deteriorate quickly, such as ataxia or anaemia, or where a daily tablet is unlikely to be taken.

None of that is a reason to change anything you have been prescribed without speaking to whoever prescribed it. It is a reason to know the conversation exists, because a great many people are still told injections are the only option available to them.

And if a B12 deficiency turns out to be real, there is a test for why. The section above names pernicious anaemia, the autoimmune attack on the stomach lining, as the most common reason absorption fails. Intrinsic factor antibodies are the test for it. A positive result is close to conclusive and it changes things, because it means the problem is permanent and that it travels with other autoimmune conditions. It does not automatically mean injections, for the reason described just above, but it does mean replacement of one kind or another for life. A negative does not rule it out, since the test misses roughly half of cases. It is not something we sell. A GP will arrange it, and private labs have it.

What changes when these are low

Tiredness, breathlessness on exertion, poor concentration and low mood are common to all three. They are also common to iron deficiency, to an underactive thyroid, and to simply not sleeping enough. That overlap is the entire argument for measuring rather than guessing which one it is, and it is why these markers sit alongside thyroid and iron on the same panel rather than being sold in isolation.

Beyond the shared picture, each has something of its own. Vitamin D deficiency at the severe end affects bone specifically: rickets in children, osteomalacia in adults, and a contribution to osteoporosis over time. B12 adds the neurological picture described above, which is the one to catch early.